Lung cancer is a major cause of death in the United States and there is a clear need for additional and better therapeutic approaches to its treatment. It is clear that there are multiple oncogenotypes that cause lung cancer and that these different genotypes respond differently to currently approved therapeutics. However, the molecular basis of these differences are not understood completely and a systematic approach to identifying most of the functional differences in cells derived from tumors having different oncogenotypes has not been carried out. The work we propose will provide a systematic study of differential sensitivities of cells with different oncogenic mutations to the loss of function of each human gene, revealing pathways that contain novel drug targets. It will be conducted on a scale large enough to distinguish genetic weaknesses that are specific for a single cell line from potential therapeutic targets that are present in all cells sharing the same set of oncogenic genetic changes. In parallel, we will use high-throughput screening to determine if any of a set of 200,000 compounds, including a subset currently approved for use in humans, has therapeutic benefit in one or more subsets of genetically distinct lung cancer. Compounds that have such activity will be tested for activity against human tumors in explant models. This will provide a potential fast-track for novel therapies as well as a method for detecting tumor specific vulnerabilities that are resistant to detection by RNAi strategies. A database of the combined functional genomic and compound screening results will be made available to the public by linking it to caGRID and uploading compound structures and linked biological data to PubChem.